Setiawan, Rizal (2026) Komparasi model You Only Look Once (YOLO) dan Convolutional Neural Network (CNN) dalam klasifikasi jenis kendaraan. Sarjana thesis, UIN Sunan Gunung Djati Bandung.
Full text not available from this repository. (Request a copy)Abstract
INDONESIA: Perkembangan kecerdasan buatan dan deep learning telah mendorong kemajuan signifikan dalam bidang pengolahan citra digital, khususnya pada tugas klasifikasi objek. Salah satu penerapannya adalah klasifikasi jenis kendaraan untuk mendukung sistem transportasi cerdas, seperti pengawasan lalu lintas dan manajemen berbasis kamera. Penelitian ini bertujuan untuk membandingkan kinerja model Convolutional Neural Network (CNN) berbasis arsitektur ResNet50 dan model You Only Look Once versi 8 (YOLOv8) dalam mengklasifikasikan jenis kendaraan, yaitu mobil, sepeda motor, truk, dan bus. Dataset yang digunakan berupa citra kendaraan dari sumber terbuka yang telah melalui tahap preprocessing, augmentation data, serta pembagian data pelatihan dan pengujian. Metodologi penelitian mengacu pada kerangka kerja CRISP-DM yang mencakup tahap pemahaman bisnis, pemahaman data, persiapan data, pemodelan, dan evaluasi. Evaluasi kinerja dilakukan menggunakan metrik accuracy, precision, recall, F1-score, serta waktu inferensi. Hasil penelitian menunjukkan bahwa model CNN berbasis ResNet50 menghasilkan akurasi dan kecepatan inferensi yang lebih baik dibandingkan YOLOv8, meskipun YOLOv8 tetap menunjukkan performa yang stabil dan kompetitif. ENGLISH: The development of artificial intelligence and deep learning has significantly advanced digital image processing, particularly in object classification tasks. One important application is vehicle type classification to support intelligent transportation systems such as traffic monitoring and camerabased management. This study aims to compare the performance of a Convolutional Neural Network (CNN) based on the ResNet50 architecture and the You Only Look Once version 8 (YOLOv8) model in classifying vehicle types, including cars, motorcycles, trucks, and buses. The dataset consists of vehicle images obtained from open-source repositories that have undergone preprocessing, data augmentation, and training–testing data splitting. The research methodology follows the CRISP-DM framework, which includes business understanding, data understanding, data preparation, modeling, and evaluation stages. Model performance is evaluated using accuracy, precision, recall, F1-score, and inference time metrics. The results indicate that the CNNbased ResNet50 model achieves higher classification accuracy and faster inference time compared to the YOLOv8 model, while YOLOv8 demonstrates stable and competitive performance across different testing scenarios.
| Item Type: | Thesis (Sarjana) |
|---|---|
| Additional Information: | tidak ada lampiran |
| Uncontrolled Keywords: | klasifikasi kendaraan; deep learning; CNN; YOLOv8; computer vision |
| Subjects: | Special Computer Methods > Computer Vision |
| Divisions: | Fakultas Sains dan Teknologi > Program Studi Teknik Informatika |
| Depositing User: | Rizal Setiawan |
| Date Deposited: | 09 Apr 2026 07:37 |
| Last Modified: | 09 Apr 2026 07:37 |
| URI: | https://digilib.uinsgd.ac.id/id/eprint/129640 |
Actions (login required)
 |
View Item |
